[
https://issues.apache.org/jira/browse/HDFS-8020?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14387894#comment-14387894
]
Kai Zheng commented on HDFS-8020:
---------------------------------
[~kaisasak],
Sure. Before that, I'd like to clarify with you first. HDFS-7717 is targeting
as a erasure coding work scheduler, which can be potentially NameNode, as
HDFS-7369 did already done by [~zhz], or separate services/tools to share the
scheduling load for NameNode. This issue together with HDFS-8021 is in DataNode
or ECWorker side, and they help to implement how to process the received
erasure coding commands from schedulers like NameNode or the ones provided by
HDFS-7717. To distribute a coding work command to DataNode, a scheduler needs
to send a BlockGroup and associated schema info in the coding command, so in
the DataNode side, it needs to restore such information back for following to
process. So HDFS-7717 needs to consider how to collect the BlockGroup and
schema information basically from NameNode, where I thought HDFS-8023 may help
if you're working on a separate scheduler tool.
> Erasure Coding: restore BlockGroup and schema info from stripping coding
> command
> --------------------------------------------------------------------------------
>
> Key: HDFS-8020
> URL: https://issues.apache.org/jira/browse/HDFS-8020
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Kai Zheng
> Assignee: Kai Zheng
>
> As a task of HDFS-7344, to process *stripping* coding commands from NameNode
> or other scheduler services/tools, we need to first be able to restore
> BlockGroup and schema information in DataNode, which will be used to
> construct coding work.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
